SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2002.06224 · NeurIPS · 2020

Top-k Training of GANs: Improving GAN Performance by Throwing Away Bad Samples

Colin Raffel, Samarth Sinha, Zhengli Zhao, Anirudh Goyal, Augustus Odena

arXiv · PDF · Open in the Atlas

Code that ran

We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

We introduce a simple (one line of code) modification to the Generative Adversarial Network (GAN) training algorithm that materially improves results with no increase in computational cost: When updating the generator parameters, we simply zero out the gradient contributions from the elements of the batch that the critic scores as 'least realistic'. Through experiments on many different GAN variants, we show that this 'top-k update' procedure is a generally applicable improvement. In order to understand the nature of the improvement, we conduct extensive analysis on a simple mixture-of-Gaussians dataset and discover several interesting phenomena. Among these is that, when gradient updates are computed using the worst-scoring batch elements, samples can actually be pushed further away from their nearest mode. We also apply our method to recent GAN variants and improve state-of-theart FID for conditional generation from 9.21 to 8.57 on CIFAR-10.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2002.06224")
get_code_for_paper("2002.06224")
have("2002.06224")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2002.06224.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2002.06224.svg)](https://syntology.ai/paper/2002.06224)

Connect an agent — have() is free.